I've thought a little more about it, and this seems to work nicely. Summarizing: QE communication can be faster than light - indeed, theoretically arbitrarily close to istantaneous. The fundamental limitation, however, comes into play when QE communication is combined with large inertial shifts. [b]Any system involving multiple QE receivers/transmitters at speed v>0 from each other deteriorates the entanglement and the speed of QE communication. If v< < c, the communication speed can still be significantly faster than light - by a factor that is to a first approximation equals 2(c/v)^2[/b]. [i]This holds even if each QE link is between points with the same velocity, as long as information has to pass through multiple links and the extremities of different links are at a relative speed greater than 0.[/i] What does this imply in practice? In terms of time travel, it should make it impossible (barring wormholes and such). The classic example of time travel with A and B sharing an inertial frame and an ansible, and C and D (colocated respectively with B and A) also sharing a (different) inertial frame and an ansible, doesn't work. Even if the QE ansibles are between pairs of points in the same inertial frame, information coming from the first QE channel, must have changed its frame of reference to enter the second QE channel, and so it slows down. In some sense, when information comes out of a QE link, even when made "classical", it carries a "memory" of having traveled through the QE link that will interfere with any other QE link it is shoved into, at least for some time (the longer the links and the larger the inertial difference, the longer the memory). In terms of the EP universe, it changes almost nothing if we handwave the issues involving Pandora gates (which involve a lot of handwaving anyways). As long as you do not bring quantum farcasters on board of anything faster than 300Km/s, you can have QE communications between any pair of points from the Vulcanoids to outer edge of the Kuiper belt with a round-trip delay of a fraction of a second - comparable to that in today's phone talks. Note that virtually any habitat orbiting the Sun or a lesser body in the system has speeds of 100Km/s less; which means that anything traveling with substantial gravitational assists will also have speeds in approximately that range (e.g. the Earth/Mars shuttle travels with a peak speed of a few dozen Km/s). With fast couriers reaching half percent the speed of light (1500Km/s) things get a little trickier, meaning delays of up to a dozen seconds between points within the Kuiper belt. This involves not only communication *between* couriers, but any communication that is mediated by a long range QE link on a courier. This means that couriers traveling at top speed try to avoid using long range QE communications or risk temporary disruption of information in any communication network they are involved in (or, more accurately, delays of up to a dozen seconds). Note that at ranges that are not system-spanning, even for couriers these delays can be quite small (we are talking of 50 milliseconds round-trip delay between points at the same distance the Earth is from the Sun, 1 AU).